There is a gender "gap" in ultra running, but that might be OK.

If you look around the startline at a 100 mile ultra marathon you'll probably notice more men than women. That is of course if you can tell the difference, they all look pretty much the same when wrapped up in compression gear and have bottles strapped to their chests. Beards seem to have gone out of fashion. 

It's probably best to look at the Ms and Fs on the start list if you want to gauge how many of each group there are in the race. If you do then in pretty much any ultra in the world there will be many more Ms than Fs. The split of Ms and Fs in the population at large is 50/50, why is it not the same at the start line of a 100 mile mountain race?

What is the explanation?

1 - This discrepancy part of a fiendish and deliberate plan to exclude women from competitive ultra events?

2 - Could it be explained by a more subtle process of social conditioning where women are gradually told over a lifetime that they do not belong on start lines like this?

3 - Or could it be explained by choice? Could Men and Women be inherently different in ways that ultimately less of them chose to be here?

Perhaps bits of all three?

What happens when you suggest one of these on facebook or twitter?

Say 1 - Race directors will take exception and defend themselves. This was a valid explanation a number of years ago.

Say 2 - Hearts and thumbs. So many hearts and thumbs.

Say 3 - Well...


I am a woman. I guess I should just give up then

I am a woman. I guess I should just give up then




I am going to try and put it on the table, with a blog, since it gets swiped off the table every time someone says it on a facebook discussion. I believe that the gender gap we see in ultras is partially driven by biology, I don't believe it will ever be a 50/50 participation rate and I don't think this is necessarily a "problem"

You may disagree. That's fine. I am open to the possibility that biology plays no part, but I don't think the evidence is there to say this with certainty, and so I don't believe people who suggest it should be lumped in with the flat-earthers in a torrent of virtue signalling online. The evidence to rule out biology completely is impossible to get, but absence of evidence is not evidence of absence. I don't think it should be discarded as a possible explanation."

First things first. Let me start with a bit of n=1 virtue signalling of my own.

--start of virtue signal--

Between 2007 and 2015 I ran a bunch of ultras. Dunno how many. Probably more than 50. I've done some of the "classics" - Spartathlon 3 times, Badwater, UTMB, Grand Union Canal Race. Some of the lesser known like the Vol State Race and the Northburn 100. In 2011 I ran the Los Angeles to New York footrace.

It was the day before the start of the Vol State race that my wife called to tell me she was pregnant, with twins. I was in Tennessee preparing to run 500k unsupported across a state. She was in New Zealand. I wasn't going to see her for a week and in that week I'd be trying to shlep across a state in really uncomfortable humidity.

I changed that day. Ultra running no longer felt so important. All I could think about were those babies. Since their birth nearly two and a half years ago I have not run an ultra marathon. If you told me before that day in Tennessee that I would not run an ultra for 2.5 years after their birth I would have laughed, or perhaps been quite sad about the prospect. I'd have been shocked to have been told I'd not run another ultra in 2.5 years after but I am more shocked by the fact that I haven't wanted to run one either.

The kids have taken over my life. I have moved jobs to spend more time with them. All my weekends are with them. I get them up every morning and put them to bed every evening. The time I used to spend running I am now spending changing nappies, making food, reading stories, pretending to be a jellyfish, holding a crying child in the night and letting them jump on the bed when Mum's not looking. What was the time I'd be running 60+ mile weeks and running some event most weekends is now totally on them. And I love it. They have been alive for about 800 days now. Wracking my brain to think of how days have gone that I've not seen them. I think it's 3. 

I have in the past year managed to train for two marathons, getting up to about 50 mile weeks. These 50 miles are rarely done on the kids time. I leave the house at 4.30am to get 10 miles done before they are awake at 6. I did my long runs on Tuesday evenings, 20 miles from about 7.30-10pm, feeling pretty tired at the end as my bed time is 9 nowadays. I might occasionally get out for a weekend run while they nap but I put the kids and family first. I will cancel a run if there is any issue and have had to pull out of a few half marathons when the kids are sick. Sometimes I'm just too tired to run after being up all night. Sometimes I don't run if I think I need to be at 100% for the kids.

But I managed to get the training in. Every hour spent on facebook is a missed opportunity to go and jog. And now later in the year I plan on running a 100 miler. Preparation for this is different. I reckon I will be fit if I carry on with the 50 mile weeks. I often think ahead to the next race, mentally rehearsing what I am going to do, but this time it's different. The "what to do when I feel sleepy or have a sore leg or am thirsty" doesn't cross my mind anymore. Now it's "how am I going to feel 12 hours in when I'm sore and sleepy and all I can think about is putting the kids to bed?"

Like I said, my brain changed that day in Tennessee. As far as ultras are concerned I want them less, and now I am going to do one the mental barriers are different.

--end of virtue signal--

OK, a bit more

--start of more virtue signalling--

My twins are boy and girl (I said Boy first obviously because he came out first). I am well aware of the stereotyping that goes on. We want them both to wear nice bright clothes, they like bright colours, but most shops sell grimy looking greys and blues with diggers and dinosaurs for boys and the sparkly princess purple bubbles filly bright stuff for girls. We often put him in a dress, because he likes it, lets face it, wearing just a dress and a nappy in the summer sounds pretty comfortable to me. Haven't tried at work yet because of "society".

I'd love for my kids to share my interests. I'd love it if they got into running and statistics/data science (my job). These are two  male skewed activities, I have no desire to push the boy into one of them any more than the girl. I'd hate for either of them to get excited about cars. They already know "Mummy's car is blue and Daddy's car is red". That's the extent of my knowledge of cars. I have nothing else to teach them on that. 

--virtue signal ends--

Let's leave biology on the table

Studies are ten a penny demonstrating that "social conditioning" is associated with the social conditions we see.  Stories of women being marginalised when getting into sports (and other things) are rife too. I am sure every woman has a story of how she's been told she shouldn't be doing [insert manly activity].

We can observe the outcomes easily enough. Men get paid more, go to prison more, race motorbikes more and women are more concerned with their appearance, take caring and teaching jobs more and spend more time with their own kids.

How much of this is the conditioning? How much is the biology? Is this a case of identical brains been given two different treatments over a lifetime and ending up with the behaviours we see? Is at least some of it a case of those brains being slightly different to start off with, reflecting somehow the different physical biological roles men and women will later play in life to propagate the species?

Is that last sentence such an outrageous thing to believe?

Conditioning (and/or/not/via) biology is one of those things that is really hard to know. You could take thousands of new born babies (or better still newly pregnant couples) and stick them in a lab for 30 years, make sure that regardless of the gender of the child is treated to the same upbringing. They get the same amount of diggers and tiaras. The girls get pushed into racing bikes and wrestling just as much as the boys. The boys get told they look pretty (my boy loves to wear a dress and prance around saying "I'm beautiful") and obsess about their appearance just as much as the girls do.

Then at the end of the 30 years we'll see. In terms of hobbies is there an equal gender split on boxing, football, crafts, cooking, writing and other pursuits. If there are then we can safely say that 100% of the differences we see in these gaps today is caused by that conditioning and 0% caused by biology.

But we can't do this type of study. It's impossible to separate. It may well be the case that all of the gap is caused by conditioning but at this point we don't have the evidence (nor are we able to get the evidence) to confirm it. Or rather we can not reject the hypothesis that the biological component of this difference is 0. I think it's reasonable to believe it's greater than zero. 

So my prediction is - "Some of the gender gap in ultra marathon participation is due to biology"

OK, you may not agree. But if you are open to the potential nonzeroness of biology's impact on a persons desire to run 100 miles in the mountains then please go ahead and suffer my ability to explain some stats stuff.

Mind the gap

When people talk about a "gap" they imagine a distribution like this. That when we say "men are more thingy than women" we are talking about two completely separate groups of people who to not overlap at all. This is the "gap instinct" I recommend reading Factfullness to get a really good view of how stats are abused.

big gap.PNG


In actual fact when we say that men are more "thingy" than women we are really saying "the distribution between the thingyness of men is skewed slighly higher than for women". Doesn't roll off the tongue. Isn't so easy to express as a single digit and isn't as convenient for blaming someone or something. 

So here is what a couple of normal distributions of thingyness, where men have an average of 51 and women have an average of 50. (variance 10).

distributon of thingyness.PNG


You can only just tell them apart! You could go through your life not knowing that men and women differ at all in thingyness. Most of the people you meet will have thingyness in the range of 25-75, you'll know some women who score 70 and say they are very very thingy, you may meet some men who score 30 and say "I don't like him cos he's not very thingy". There is practically no difference between these two groups, none that you'll really notice in your every day life.


extreme thingyness.PNG


When you zoom into the extreme, to the numbers of people who score 90 or higher for thingyness, what do you see here? Way more men than women. About 0.04% of the female population score above 90 whereas about 0.06% of the men get there. So (assuming there are the same number of men and women to start with) the ratio of men to women in the 90+ score the split is 60% and 40%. That's a massive "gender gap" on what is only a slight and almost unnoticeable skew in normal life.

If we allowed for men to have an average thingyness of 52, so still only 4% higher on average and still pretty much unnoticeable in normal life this gap at the 90+ fringe extends to 69/31. 

See where I'm going with this?

If "Thingyness" was some sort of amalgamation of attributes that are required for running 100 miles in the mountains. Risk taking, narcisim, desire for kudos, confidence, over-confidence, lack of empathy or connection to others, response to dopamine and adrenaline etc and that only those who score 90+ will have any desire to run 100 miles in the mountains then you get this gender gap without much of of a difference between normal people. An almost unnoticeable difference between men and women manifests itself as a much bigger difference at the extremes.

So my overall point, you could still observe massive gender gaps in the extreme ends of sports without their being much underlying differences between the two sexes. Averages are easy to calculate and put in articles but they mask so much information. Think about what you are doing when you calculate an "average" of 10000 people. You are creating one figure and then discarding 10000 bits of information. An average is not a great descriptor of how different some things are.

OK. Second thing, let's take a diversion via some bullet ridden planes

Abraham Wald was a statistician employed in the second world war to help with the design of planes. They needed armour, but armour is heavy so not too much. He looked at a sample of planes that had been in action and looked at the distribution of bullet holes. Most were in the wings and fusilage, very few bullet holes were around the engine. Where should we put more armour? Where the bullet holes are of course.


Wald had a great insight, he asked "what about the planes that don't come back?" Any decision using this sample was going to be prone to survival bias, the tendancy to only look at the things that have already passed some "test", in this case, not getting shot down in battle. Wald decided to put the armour around the engine, since there would have been plenty of planes with holes in the engine, it's just that we can not see them because they are on the floor behind enemy lines.

And so, back to ultras. Here we are asking "why are less women not running ultras" and who are we asking? People who run ultras! We are not asking the most important group of people, those who don't run ultras. I think there is little to be gained in asking only those who have "survived" in this way what the barriers are to survival. 

wald pie.PNG


We should be asking the 99.8% of men and 99.9% of women out there who are not at the start-line of a 100 mile mountain race why they are not. There is little to be gained in sharing articles to the blue people.

Ultimately what I find most difficult about talks of "gaps" is that it's often an individual saying "this percentage is wrong, I want it to be this other percentage" and then offers "society" as the explanation to the difference, a faceless villain against whom we can thrash our virtue signalling clubs against on social media. It might be the result of free will, and it might be the case that the free will of those 99.9% of people who don't do a thing might just be a bit different from the 0.1% of people who do. "I want people to be more like me, the fact they aren't is because of "society," I am going to fix it via top-down compulsion".

So here I am arguing that there may not be a problem at all, and so am nervous about top-down "fixes" to this problem. I am going to do my bit in my own interactions though. I will keep on putting my kids first and my running way down. I will spend as much time with both of them, try to inspire them into sport and activities but ultimately taking a lead from them as to what they want to do. I will do the occasional ultra, I will talk all day to anyone about ultras if they want to. I'm not going to stress about who hasn't made the same choices as me. 

Further Reading.

I haven't scattered links to "peer reviewed" papers supporting my points. It's more an attempt at a logical flow of possibility. It is shaped by a lot of books I have read and a few are here.


Yes! He is the cousin of Ali G, but also has written a lot about empathy and how brains relate to others. Including a view then Men are (slightly hgiher skewed) towards "systemising"


The brilliant late Hans Rosling talks about why wleap to the wrong conclusions about statistics, particularly when we talk about "gaps"


Great chapter on the Wald story


Running and Stuff
By James Adams

Story about a big girl who just shits himself and cries